Quantitative LC-MS/MS method validations and comprehensive extraction for pinostrobin and cardamonin from Boesenbergia rotunda (L.) Mansf. guided by response surface methodology

The phytochemical extraction of the flavonoids cardamonin and pinostrobin from Boesenbergia rotunda is hindered by their low aqueous solubility and the matrix's susceptibility to water-induced starch gelatinization. This study investigates in silico optimal extraction parameters, including temperature, pressure, and ethanolic mole fraction, to maximize yields while preventing matrix gelatinization and artifactual isomerization. A Box-Behnken response surface methodology (RSM) was coupled with molecular dynamics (MD) simulations to screen for the lowest solvation free energy. This theoretical model was experimentally verified using five extraction methods: maceration, soxhlet, microwave-assisted, ultrasonication, and autoclave extraction. Furthermore, analyte recovery was quantified through rigorous method validation for liquid chromatography-tandem mass spectrometry (LC-MS/MS). While the predictive model identified high-temperature, water-rich solvents as optimal, experimental trials proved these conditions triggered severe starch gelatinization, creating a viscous mass-transfer barrier that trapped the analytes. Moreover, extraction techniques applying continuous thermal energy forced the artifactual cyclization of cardamonin into pinostrobin. Ultimately, absolute ethanol was empirically proven to be the superior solvent; its low dielectric constant effectively solubilized the targets while completely bypassing the gelatinization barrier to safely maximize yields. Finally, the LC-MS/MS method was confirmed to be highly accurate, reproducible, and within acceptable ranges for these evaluations.